TY - GEN
T1 - A green isolated word sign language recognition system based on geodesic metric space
AU - Fakhfakh, Sana
AU - Jemaa, Yousra Ben
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - This paper presents a new isolated word sign language (IWSL) recognition strategy to attempt an ecological and friendly solution based on the deaf brain pipeline. Two major steps are proposed in this research field.The first one is based on hands shape-motion extraction and the second is based on its analysis without introducing deep learning frameworks (DLF) to carry out less data, less time execution and less carbon footprint consumption.So, we focus on developing an energy-saving system offering a green IWSL recognition system based on geodesic distance as an elastic measure while preserving the same DLF performance. Results indicate that the new IWSL recognition system is a promising solution as it trades off energy and accuracy compared with existing systems.
AB - This paper presents a new isolated word sign language (IWSL) recognition strategy to attempt an ecological and friendly solution based on the deaf brain pipeline. Two major steps are proposed in this research field.The first one is based on hands shape-motion extraction and the second is based on its analysis without introducing deep learning frameworks (DLF) to carry out less data, less time execution and less carbon footprint consumption.So, we focus on developing an energy-saving system offering a green IWSL recognition system based on geodesic distance as an elastic measure while preserving the same DLF performance. Results indicate that the new IWSL recognition system is a promising solution as it trades off energy and accuracy compared with existing systems.
KW - complexity redaction
KW - geodesic distance
KW - Isolated word sign language recognition
KW - performance analysis
UR - http://www.scopus.com/inward/record.url?scp=85181047967&partnerID=8YFLogxK
U2 - 10.1109/CW58918.2023.00049
DO - 10.1109/CW58918.2023.00049
M3 - Conference contribution
AN - SCOPUS:85181047967
T3 - Proceedings - 2023 International Conference on Cyberworlds, CW 2023
SP - 274
EP - 281
BT - Proceedings - 2023 International Conference on Cyberworlds, CW 2023
A2 - Amara, Najoua Essoukri Ben
A2 - Sourin, Alexei
A2 - Sourina, Olga
A2 - Rosenberger, Christophe
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 22nd International Conference on Cyberworlds, CW 2023
Y2 - 3 October 2023 through 5 October 2023
ER -
